My logroll code, please critique


#1

My goal is to keep 10 files each at 100Mb. Please critique and suggest
any corrections I need to make. Thank you!

#!/usr/bin/ruby -w -W 2

require ‘ftools’
require ‘fileutils’

=begin
#-----------------------------
File::LOCK_SH # shared lock (for reading)
File::LOCK_EX # exclusive lock (for writing)
File::LOCK_NB # non-blocking request
File::LOCK_UN # free lock
#-----------------------------
=end

###-- 10 files at 100Mb before rolling begins —##

FSIZE = 104860000

Dir.chdir("/cygdrive/c/temp/log")
or raise StandardError, “Change dir failed to ~aevrlog!”

unless File.exists?("/cygdrive/c/temp/log/zipped")
File.makedirs(“zipped”)
or raise StandardError, “Make dir failed to ~aevrlog/zipped!”
end

array1 = Dir.glob(“dev*.log”)
array2 = Dir.glob(“mon*.log”)
array3 = Dir.glob(“fer*_se*.out”)
array4 = Dir.glob(“fer*_se*.log”)
array5 = Dir.glob(“fer*_in*.log”)

if (array1.length >= 1)
array1.each { |file|

if (File.size(file) > FSIZE)
    `/usr/bin/gzip "#{file}"`
    FileUtils.mv("#{file}.gz","/cygdrive/c/temp/log/zipped")
    FileUtils.touch("development.log")
end
}

end

Dir.chdir("/cygdrive/c/temp/log/zipped")
or raise StandardError, “Change dir failed to ~aevrlog!”

array1_1 = Dir["/cygdrive/c/temp/log/zipped/dev*.gz"]

if (array1_1.length >= 1)
    if File.exists?("development.log.11.gz")
        FileUtils.rm_r

Dir.glob("/cygdrive/c/temp/log/zipped/dev*11.gz")
end
if File.exists?(“development.log.10.gz”)
File.copy(“development.log.10.gz”, “development.log.11.gz”)
end
if File.exists?(“development.log.8.gz”)
File.copy(“development.log.8.gz”, “development.log.9.gz”)
end
if File.exists?(“development.log.7.gz”)
File.copy(“development.log.7.gz”, “development.log.8.gz”)
end
if File.exists?(“development.log.6.gz”)
File.copy(“development.log.6.gz”, “development.log.7.gz”)
end
if File.exists?(“development.log.5.gz”)
File.copy(“development.log.5.gz”, “development.log.6.gz”)
end
if File.exists?(“development.log.4.gz”)
File.copy(“development.log.4.gz”, “development.log.5.gz”)
end
if File.exists?(“development.log.3.gz”)
File.copy(“development.log.3.gz”, “development.log.4.gz”)
end
if File.exists?(“development.log.2.gz”)
File.copy(“development.log.2.gz”, “development.log.3.gz”)
end
if File.exists?(“development.log.1.gz”)
File.copy(“development.log.1.gz”, “development.log.2.gz”)
end
if File.exists?(“development.log.gz”)
File.copy(“development.log.gz”, “development.log.1.gz”)
end
end


#2

Derek S. removed_email_address@domain.invalid wrote:

    if File.exists?("development.log.10.gz")
        File.copy("development.log.10.gz", "development.log.11.gz")
    end
    if File.exists?("development.log.8.gz")
        File.copy("development.log.8.gz", "development.log.9.gz")
    end
    # etc, over and over and over and over

You forgot log.9, and the very fact that this can happen, along with the
patterned repetition here, would appear to be a Bad Smell (as in
“Refactoring”). Why 30 hard-coded lines when a 5-line loop will do? m.


#3

On 05.05.2009 03:52, matt neuburg wrote:

You forgot log.9, and the very fact that this can happen, along with the
patterned repetition here, would appear to be a Bad Smell (as in
“Refactoring”). Why 30 hard-coded lines when a 5-line loop will do? m.

Another thing that struck me is the usage of File.copy. File.rename
would be much more efficient plus it would not change the last
modification time of the file.

Kind regards

robert


#4

On Tue, May 5, 2009 at 2:55 AM, Tom C. removed_email_address@domain.invalid wrote:

backchannel it, if you wish). I could definitely USE this!

Take a look at the “logging” gem. The RollingFile appender will do
exactly this – rolling files based on size or age. Moreover, it will
safely roll the files even if multiple processes are writing to the
same file.

rubyforge.org/projects/logging

Blessings,
TwP


#5

Derek S. wrote:

=begin

array1 = Dir.glob(“dev*.log”)
FileUtils.mv("#{file}.gz","/cygdrive/c/temp/log/zipped")
if (array1_1.length >= 1)
if File.exists?(“development.log.7.gz”)
end
File.copy(“development.log.gz”, “development.log.1.gz”)
end
end

Derek, I’m interested in what your trying to do here, and in the
comments it’s elicited. Could you share you revised code, when it’s
ready (or backchannel it, if you wish). I could definitely USE this!

Tom


#6

On May 5, 2009, at 15:44, Derek S. wrote:

Will anyone share good sample code for logging?
Thanks to all that replied, but Matt will you show me this loop you
referred to?

I use SyslogLogger (which adapts Syslog to a Logger API) and newsyslog
to rotate logs.


#7

Tim P. wrote:

On Tue, May 5, 2009 at 2:55 AM, Tom C. removed_email_address@domain.invalid wrote:

backchannel it, if you wish). I could definitely USE this!

Take a look at the “logging” gem. The RollingFile appender will do
exactly this – rolling files based on size or age. Moreover, it will
safely roll the files even if multiple processes are writing to the
same file.

rubyforge.org/projects/logging

Blessings,
TwP

I did look into this gem, but the documentation was weak and I tried to
work it, but could not and failed so I wrote my own since I am new to
Ruby. A good learning exp!..yucky Perl, ugly code is not for me!

Will anyone share good sample code for logging?
Thanks to all that replied, but Matt will you show me this loop you
referred to?

Thank you!


#8

Derek S. wrote:

Will anyone share good sample code for logging?

There’s the logger.rb in the ruby standard library (or did someone
mention that already?). It handles rotation, log levels, but perhaps not
multiprocess safety.


#9

Joel VanderWerf wrote:

Derek S. wrote:

Will anyone share good sample code for logging?

There’s the logger.rb in the ruby standard library (or did someone
mention that already?). It handles rotation, log levels, but perhaps not
multiprocess safety.

Here is my updated code, but how do I look at the documentation from the
CLI for logger, logging or any gem? For example, if I want to look at
Perl docs I use perldoc -f “builtin function” or perldoc -q “F.A.Q.”
Finally, what about a central logging block for when at anytime ‘raise
StandardError’ is called, the script will log to one logfile. Will
anyone suggest code or how to do this?

thank you!

#!/usr/bin/ruby -w -W 2

require ‘ftools’
require ‘fileutils’

=begin
#-----------------------------
File::LOCK_SH # shared lock (for reading)
File::LOCK_EX # exclusive lock (for writing)
File::LOCK_NB # non-blocking request
File::LOCK_UN # free lock
#-----------------------------
=end

###-- 10 files at 100Mb before rolling begins —##

FSIZE = 104860000

Dir.chdir("/cygdrive/c/temp/log")
or raise StandardError, “Change dir failed to ~aevrlog!”

unless File.exists?("/cygdrive/c/temp/log/zipped")
File.makedirs(“zipped”)
or raise StandardError, “Make dir failed to ~aevrlog/zipped!”
end

array1 = Dir.glob(“dev*.log”)
array2 = Dir.glob(“mon*.log”)
array3 = Dir.glob(“fer*_se*.out”)
array4 = Dir.glob(“fer*_se*.log”)
array5 = Dir.glob(“fer*_in*.log”)

if (array1.length >= 1)
array1.each { |file|

if (File.size(file) > FSIZE)
    File.open(file, "r+") do |f|
        if (f.flock(File::LOCK_SH)) == 0
            `/usr/bin/gzip "#{file}"`
             FileUtils.mv("#{file}.gz","/cygdrive/c/temp/log/zipped")
        end
    end
    FileUtils.touch("development.log")
end
}

end

Dir.chdir("/cygdrive/c/temp/log/zipped")
or raise StandardError, “Change dir failed to ~aevrlog!”

array1_1 = Dir["/cygdrive/c/temp/log/zipped/dev*.gz"]

if (array1_1.length >= 1)
    if File.exists?("development.log.11.gz")
        FileUtils.rm_r 

Dir.glob("/cygdrive/c/temp/log/zipped/dev*11.gz")
end
if File.exists?(“development.log.10.gz”)
File.rename(“development.log.10.gz”,
“development.log.11.gz”)
end
if File.exists?(“development.log.9.gz”)
File.rename(“development.log.9.gz”, “development.log.10.gz”)
end
if File.exists?(“development.log.8.gz”)
File.rename(“development.log.8.gz”, “development.log.9.gz”)
end
if File.exists?(“development.log.7.gz”)
File.rename(“development.log.7.gz”, “development.log.8.gz”)
end
if File.exists?(“development.log.6.gz”)
File.rename(“development.log.6.gz”, “development.log.7.gz”)
end
if File.exists?(“development.log.5.gz”)
File.rename(“development.log.5.gz”, “development.log.6.gz”)
end
if File.exists?(“development.log.4.gz”)
File.rename(“development.log.4.gz”, “development.log.5.gz”)
end
if File.exists?(“development.log.3.gz”)
File.rename(“development.log.3.gz”, “development.log.4.gz”)
end
if File.exists?(“development.log.2.gz”)
File.rename(“development.log.2.gz”, “development.log.3.gz”)
end
if File.exists?(“development.log.1.gz”)
File.rename(“development.log.1.gz”, “development.log.2.gz”)
end
if File.exists?(“development.log.gz”)
File.rename(“development.log.gz”, “development.log.1.gz”)
end
end


#10

Derek S. wrote:

Here is my updated code, but how do I look at the documentation from the
CLI for logger, logging or any gem? For example, if I want to look at
Perl docs I use perldoc -f “builtin function” or perldoc -q “F.A.Q.”

http://www.ruby-doc.org/stdlib/

Look in the left panel for “logger”.

This site is still pretty useful (though I don’t know if it is
frequently updated).


#11

Derek S. wrote:

Here is my updated code, but how do I look at the documentation from the
CLI for logger, logging or any gem? For example, if I want to look at
Perl docs I use perldoc -f “builtin function” or perldoc -q “F.A.Q.”
Finally, what about a central logging block for when at anytime ‘raise
StandardError’ is called, the script will log to one logfile. Will
anyone suggest code or how to do this?

More on finding documentation -

I can’t help you if you’re on a Winxxxx system, but on my Kubuntu Linux
OS I can find useful documentation for the “logger” standard library
member (and others) by issuing at my CLI -

which logger

I get a pile of locations on the filesys where “logger” appears. Many
are irrelevant, but this one is not:

/usr/share/doc/ruby1.8-examples/examples/logger

There I find 3 ruby code files illustrating library use. Nice.

(I didn’t know about this until I set out to find an answer to your
question, so thanks to you for that.)

t.

Tom C., MS MA, LMHC - Private practice Psychotherapist
Bellingham, Washington, U.S.A: (360) 920-1226
<< removed_email_address@domain.invalid >> (email)
<< TomCloyd.com >> (website)
<< sleightmind.wordpress.com >> (mental health weblog)

#12

Derek S. wrote:

Here is my updated code, but how do I look at the documentation from the
CLI for logger, logging or any gem? For example, if I want to look at
Perl docs I use perldoc -f “builtin function” or perldoc -q “F.A.Q.”
Finally, what about a central logging block for when at anytime ‘raise
StandardError’ is called, the script will log to one logfile. Will
anyone suggest code or how to do this?

Google “ruby logger” and you get right to the RDoc documentation -
http://www.ruby-doc.org/stdlib/libdoc/logger/rdoc/index.html

Access the “logger” class and you’ll get some very good documentation.

Re: errors - the standard error stream goes to STDERR, a constant ruby
automatically creates. Normally, this goes to STDOUT - the standard
output stream, assigned to your screen, but you can reassign it:

$stderr = File.new( ‘myerrs.txt’,‘w’ )

Hope this helps…

t.

Tom C., MS MA, LMHC - Private practice Psychotherapist
Bellingham, Washington, U.S.A: (360) 920-1226
<< removed_email_address@domain.invalid >> (email)
<< TomCloyd.com >> (website)
<< sleightmind.wordpress.com >> (mental health weblog)